RISE Seminar: Learning to Solve Combinatorial Optimization Problems with Applications to Systems and Chip Design, a talk by Azalia Mirhoseini
May 14, 2020
This week, we are very excited to host Azalia Mirhoseini from Google Brain. Azalia will tell us about reinforcement learning in systems and chip design.
Speaker: Azalia Mirhoseini (Google Brain)
Title: Learning to Solve Combinatorial Optimization Problems with Applications to Systems and Chip Design
Date & Time: Thursday, May 14 2020, 4-5pm
Zoom Webinar: https://berkeley.
Abstract: In the past decade, computer systems and chips
have played a key role in the success of AI. Our vision is to use AI to transform the way systems and chips are designed. Many core problems in systems and hardware design are combinatorial optimization or decision making tasks with state and action sizes that are orders of magnitude larger than common AI benchmarks in robotics and games. In this talk, I will go over some of our research on tackling such optimization problems. First, I will talk about our work on deep reinforcement learning models that learn to do computational resource allocation, a combinatorial optimization problem that repeatedly appears in systems. Our method is end-to-end and abstracts away the complexity of the underlying optimization space; the RL agent learns the implicit tradeoffs between computation and communication of the underlying resources and optimizes the allocation using only the true reward function (e.g., the runtime of the generated allocation). Then, I will talk about our work on generalizable graph clustering and partitioning with unsupervised learning. Finally, I will discuss our recent work on optimizing chip placement with reinforcement learning. Our approach has the ability to learn from past experience and improve over time. The placement policy can generalize to unseen blocks. Our objective is to minimize PPA (power, performance, and area), and we show that, in under 6 hours, our method can generate placements that are superhuman or comparable on modern accelerator chips, whereas existing baselines require human experts in the loop and can take several weeks.
Bio: Azalia is a Senior Research Scientist at Google Brain and an advisor at Cmorq. She is the co-founder/lead of the Machine Learning for Systems Moonshot at Brain where they focus on deep reinforcement learning based approaches to solve problems in computer systems and metalearning. Azalia has a Ph.D. in Electrical and Computer Engineering from Rice University. She has received a number of awards, including the MIT Technology Review 35 Under 35 Award, the Best Ph.D. Thesis Award at Rice University, and a Gold Medal in the National Math Olympiad in Iran.